An Efficient Forward Recovery Checkpointing Scheme in Distributed Systems

Abstract

In this paper, we present a cost-effective forward recovery checkpointing scheme (FRCS)for distributed systems without a spare module, where a job is executed simultaneously on two processors (or processing modules). For those schemes with a spare module, the time required for initiating the spare module is long and the cost is high for an extra module. Mathematical models have been developed to demonstrate that our scheme is always better than the traditional rollback scheme without any spare module and is better than other schemes with a spare module in terms of number of processing modules required and the total execution time (cost-effectiveness).